Neural network versus econometric models in forecasting inflation

Artificial neural network modelling has recently attracted much attention as a new technique for estimation and forecasting in economics and finance. The chief advantages of this new approach are that such models can usually find a solution for very complex problems, and that they are free from the assumption of linearity that is often adopted to make the traditional methods tractable. In this paper we compare the performance of Back‐Propagation Artificial Neural Network (BPN) models with the traditional econometric approaches to forecasting the inflation rate. Of the traditional econometric models we use a structural reduced‐form model, an ARIMA model, a vector autoregressive model, and a Bayesian vector autoregression model. We compare each econometric model with a hybrid BPN model which uses the same set of variables. Dynamic forecasts are compared for three different horizons: one, three and twelve months ahead. Root mean squared errors and mean absolute errors are used to compare quality of forecasts. The results show the hybrid BPN models are able to forecast as well as all the traditional econometric methods, and to outperform them in some cases. Copyright © 2000 John Wiley & Sons, Ltd.     

Forecasting Time Series with Trading Day Variation

Some levels of economic activity change over the days of the week. Also, the composition of the calendar changes over the years so that a particular month contains a different configuration of days of the week each year. The effects of the changing composition of the calendar upon a monthly time series is called trading day variation. This paper discusses one way to model trading day variation in monthly time series and how this model can be used to obtain improved forecasts over univariate ARIMA models. The ideas are illustrated on an actual data set.     

ARIMA models of the price level: An assessment of the multilevel adaptive learning process in the USA

This paper estimates the ARIMA processes for the observed and expected price level corresponding to the three-level adaptive expectations model proposed by Jacobs and Jones (1980). These univariate processes are then compared with the best-fit ARIMA model. The results indicate that the best-fit model for the observed price level is a restricted version of the two-level adaptive learning process specified in terms of prices, suggesting a simple adaptive rule in the inflation rate. A comparison of the time-series forecasts from the best-fit model with the mean responses to the ASA-NBER survey shows no significant difference in their accuracy. The time-series forecasts are, however, conditionally efficient. The best-fit ARIMA model for expected prices measured by the ASA-NBER consensus forecasts does not correspond to any version of the Jacobs and Jones model.     

Can cointegration‐based forecasting outperform univariate models? An application to Asian exchange rates

Conventional wisdom holds that restrictions on low‐frequency dynamics among cointegrated variables should provide more accurate short‐ to medium‐term forecasts than univariate techniques that contain no such information; even though, on standard accuracy measures, the information may not improve long‐term forecasting. But inconclusive empirical evidence is complicated by confusion about an appropriate accuracy criterion and the role of integration and cointegration in forecasting accuracy. We evaluate the short‐ and medium‐term forecasting accuracy of univariate Box–Jenkins type ARIMA techniques that imply only integration against multivariate cointegration models that contain both integration and cointegration for a system of five cointegrated Asian exchange rate time series. We use a rolling‐window technique to make multiple out of sample forecasts from one to forty steps ahead. Relative forecasting accuracy for individual exchange rates appears to be sensitive to the behaviour of the exchange rate series and the forecast horizon length. Over short horizons, ARIMA model forecasts are more accurate for series with moving‐average terms of order >1. ECMs perform better over medium‐term time horizons for series with no moving average terms. The results suggest a need to distinguish between ‘sequential’ and ‘synchronous’ forecasting ability in such comparisons. Copyright © 2002 John Wiley & Sons, Ltd.     

Forecasting with balanced state space representations of multivariate distributed lag models

A procedure for estimating state space models for multivariate distributed lag processes is described. It involves singular value decomposition techniques and yields an internally balanced state space representation which has attractive properties. Following the specifications of a forecasting competition, the approach is applied to generate ex-post forecasts for US real GNP growth rates. The forecasts of the estimated state space model are compared to those of twelve econometric models and an ARIMA model.     

Theory and practice of multivariate arma forecasting

We compare univariate and multivariate forecasts based on ARMA models. In theory we cannot do worse by using a multivariate model instead of a univariate one, but we can risk getting no improvement. Conditions for no improvements are discussed as well as cases where large improvements occur. The effect of estimated parameters is examined and found to be small granted that a good method of estimation is used. However, multivariate models could be very sensitive to structural changes. This is illustrated via an example involving monetary data, where the multivariate forecasts perform considerably worse than the univariate ones. This seems to put a limitation on the use of multivariate ARMA forecasting models.     

Forecasting European GNP data through common factor models and other procedures

In this paper we present an extensive study of annual GNP data for five European countries. We look for intercountry dependence and analyse how the different economies interact, using several univariate ARIMA and unobserved components models and a multivariate model for the GNP incorporating all the common information among the variables. We use a dynamic factor model to take account of the common dynamic structure of the variables. This common dynamic structure can be non‐stationary (i.e. common trends) or stationary (i.e. common cycles). Comparisons of the models are made in terms of the root mean square error (RMSE) for one‐step‐ahead forecasts. For this particular group of European countries, the factor model outperforms the remaining ones. Copyright © 2002 John Wiley & Sons, Ltd.     

Judgement in learning‐curve forecasting: a laboratory study

This study investigates whether human judgement can be of value to users of industrial learning curves, either alone or in conjunction with statistical models. In a laboratory setting, it compares the forecast accuracy of a statistical model and judgemental forecasts, contingent on three factors: the amount of data available prior to forecasting, the forecasting horizon, and the availability of a decision aid (projections from a fitted learning curve). The results indicate that human judgement was better than the curve forecasts overall. Despite their lack of field experience with learning curve use, 52 of the 79 subjects outperformed the curve on the set of 120 forecasts, based on mean absolute percentage error. Human performance was statistically superior to the model when few data points were available and when forecasting further into the future. These results indicate substantial potential for human judgement to improve predictive accuracy in the industrial learning‐curve context. Copyright © 1999 John Wiley & Sons, Ltd.     

Linear constraints,robust-weighting and efficient composite modeling

Recent studies have shown that composite forecasting produces superior forecasts when compared to individual forecasts. This paper extends the existing literature by employing linear constraints and robust regression techniques in composite model building. Security analysts forecasts may be improved when combined with time series forecasts for a diversified sample of 261 firms with a 1980-1982 post-sample estimation period. The mean square error of analyst forecasts may be reduced by combining analyst and univariate time series model forecasts in constrained and unconstrained ordinary least squares regression models. These reductions are very interesting when one finds that the univariate time series model forecasts do not substantially deviate from those produced by ARIMA (0,1,1) processes. Moreover, security analysts' forecast errors may be significantly reduced when constrained and unconstrained robust regression analyses are employed.     

The Effects of Disaggregation on Forecasting Nonstationary Time Series

This paper focuses on the effects of disaggregation on forecast accuracy for nonstationary time series using dynamic factor models. We compare the forecasts obtained directly from the aggregated series based on its univariate model with the aggregation of the forecasts obtained for each component of the aggregate. Within this framework (first obtain the forecasts for the component series and then aggregate the forecasts), we try two different approaches: (i) generate forecasts from the multivariate dynamic factor model and (ii) generate the forecasts from univariate models for each component of the aggregate. In this regard, we provide analytical conditions for the equality of forecasts. The results are applied to quarterly gross domestic product (GDP) data of several European countries of the euro area and to their aggregated GDP. This will be compared to the prediction obtained directly from modeling and forecasting the aggregate GDP of these European countries. In particular, we would like to check whether long‐run relationships between the levels of the components are useful for improving the forecasting accuracy of the aggregate growth rate. We will make forecasts at the country level and then pool them to obtain the forecast of the aggregate. The empirical analysis suggests that forecasts built by aggregating the country‐specific models are more accurate than forecasts constructed using the aggregated data. Copyright © 2014 John Wiley & Sons, Ltd.     

A Dynamic Factor Approach to Mortality Modeling

Longevity risk has become one of the major risks facing the insurance and pensions markets globally. The trade in longevity risk is underpinned by accurate forecasting of mortality rates. Using techniques from macroeconomic forecasting we propose a dynamic factor model of mortality that fits and forecasts age‐specific mortality rates parsimoniously. We compare the forecasting quality of this model against the Lee–Carter model and its variants. Our results show the dynamic factor model generally provides superior forecasts when applied to international mortality data. We also show that existing multifactorial models have superior fit but their forecasting performance worsens as more factors are added. The dynamic factor approach used here can potentially be further improved upon by applying an appropriate stopping rule for the number of static and dynamic factors. Copyright © 2013 John Wiley & Sons, Ltd.     

Univariate Forecasting Comparisons: The Case of the Spanish Automobile Industry

This paper investigates the forecasting ability of unobserved component models, when compared with the standard ARIMA univariate approach. A forecasting exercise is carried out with each method, using monthly time series of automobile sales in Spain. The accuracy of the different methods is assessed by comparing several measures of forecasting performance based on the out-of-sample predictions for various horizons, as well as different assumptions on the models’ parameters. Overall there seems little to choose between the methods in forecasting performance terms but the recursive unobserved component models provide greater flexibility for adaptive applications. © 1997 by John Wiley & Sons, Ltd.     

Municipal budget forecasting with multivariate ARMA models

In this paper multivariate ARMA models are applied to the problem of forecasting city budget variables. Unlike univariate time-series methods, multivariate models can use relationships among budget variables as well as relationships with economic and demographic indicators. Although available budget series are shorter than what is usually believed necessary for multivariate ARMA modelling, the forecasts seem to be of higher quality than those from univariate models.     

Forecasting for business planning: A case study of IBM product sales

This is a case study of a closely managed product. Its purpose is to determine whether time-series methods can be appropriate for business planning. By appropriate, we mean two things: whether these methods can model and estimate the special events or features that are often present in sales data; and whether they can forecast accurately enough one, two and four quarters ahead to be useful for business planning. We use two time-series methods, Box-Jenkins modeling and Holt-Winters adaptive forecasting, to obtain forecasts of shipments of a closely managed product. We show how Box-Jenkins transfer-function models can account for the special events in the data. We develop criteria for choosing a final model which differ from the usual methods and are specifically directed towards maximizing the accuracy of next-quarter, next-half-year and next-full-year forecasts. We find that the best Box-Jenkins models give forecasts which are clearly better than those obtained from Holt-Winters forecast functions, and are also better than the judgmental forecasts of IBM's own planners. In conclusion, we judge that Box-Jenkins models can be appropriate for business planning, in particular for determining at the end of the year baseline business-as-usual annual and monthly forecasts for the next year, and in mid-year for resetting the remaining monthly forecasts.     

The Effect of Estimating Parameters on Long‐Term Forecasts for Cointegrated Systems

This paper concerns Long‐term forecasts for cointegrated processes. First, it considers the case where the parameters of the model are known. The paper analytically shows that neither cointegration nor integration constraint matters in Long‐term forecasts. It is an alternative implication of Long‐term forecasts for cointegrated processes, extending the results of previous influential studies. The appropriate Mote Carlo experiment supports our analytical result. Secondly, and more importantly, it considers the case where the parameters of the model are estimated. The paper shows that accuracy of the estimation of the drift term is crucial in Long‐term forecasts. Namely, the relative accuracy of various Long‐term forecasts depends upon the relative magnitude of variances of estimators of the drift term. It further experimentally shows that in finite samples the univariate ARIMA forecast, whose drift term is estimated by the simple time average of differenced data, is better than the cointegrated system forecast, whose parameters are estimated by the well‐known Johansen's ML method. Based upon finite sample experiments, it recommends the univariate ARIMA forecast rather than the conventional cointegrated system forecast in finite samples for its practical usefulness and robustness against model misspecifications. Copyright © 2011 John Wiley & Sons, Ltd.     

Assessing the historical accuracy of regional economic forecasts

This study addresses for the first time systematic evaluation of a widely used class of forecasts, regional economic forecasts. Ex ante regional structural equation model forecasts are analysed for 19 metropolitan areas. One- to ten-quarter-ahead forecasts are considered and the seven-year sample spans a complete business cycle. Counter to previous speculation in the literature, (1) dependency on macroeconomic forecasting model inputs does not substantially erode accuracy relative to univariate extrapolative methodologies and (2) stochastic time series models do not on average, yield more accurate regional economic predictions than structural models. Similar to findings in other studies, clear preferences among extrapolative methodologies do not emerge. Most general conclusions, however, are subject to caveats based on step-length effects and region-specific effects.     

Forecasting with latent structure time series models: an application to nominal interest rates

In this paper we develop a latent structure extension of a commonly used structural time series model and use the model as a basis for forecasting. Each unobserved regime has its own unique slope and variances to describe the process generating the data, and at any given time period the model predicts a priori which regime best characterizes the data. This is accomplished by using a multinomial logit model in which the primary explanatory variable is a measure of how consistent each regime has been with recent observations. The model is especially well suited to forecasting series which are subject to frequent and/or major shocks. An application to nominal interest rates shows that the behaviour of the three‐month US Treasury bill rate is adequately explained by three regimes. The forecasting accuracy is superior to that produced by a traditional single‐regime model and a standard ARIMA model with a conditionally heteroscedastic error. Copyright © 1999 John Wiley & Sons, Ltd.     

Comparing the DSGE model with the factor model: an out‐of‐sample forecasting experiment

In this paper, we put dynamic stochastic general equilibrium DSGE forecasts in competition with factor forecasts. We focus on these two models since they represent nicely the two opposing forecasting philosophies. The DSGE model on the one hand has a strong theoretical economic background; the factor model on the other hand is mainly data‐driven. We show that incorporating a large information set using factor analysis can indeed improve the short‐horizon predictive ability, as claimed by many researchers. The micro‐founded DSGE model can provide reasonable forecasts for US inflation, especially with growing forecast horizons. To a certain extent, our results are consistent with the prevailing view that simple time series models should be used in short‐horizon forecasting and structural models should be used in long‐horizon forecasting. Our paper compares both state‐of‐the‐art data‐driven and theory‐based modelling in a rigorous manner. Copyright © 2008 John Wiley & Sons, Ltd.     

Factor forecasts for the UK

Data are now readily available for a very large number of macroeconomic variables that are potentially useful when forecasting. We argue that recent developments in the theory of dynamic factor models enable such large data sets to be summarized by relatively few estimated factors, which can then be used to improve forecast accuracy. In this paper we construct a large macroeconomic data set for the UK, with about 80 variables, model it using a dynamic factor model, and compare the resulting forecasts with those from a set of standard time‐series models. We find that just six factors are sufficient to explain 50% of the variability of all the variables in the data set. These factors, which can be shown to be related to key variables in the economy, and their use leads to considerable improvements upon standard time‐series benchmarks in terms of forecasting performance. Copyright © 2005 John Wiley & Sons, Ltd.     

Time-series forecasting of the German unemployment rate

The purpose of the paper is to investigate the accuracy of forecasts derived from univariate and multivariate time-series models. An iterative method to adjust for impact assessment in univariate ARIMA models is discussed and illustrated for the German unemployment rate. Finally, we also examine the pros and cons of the impact assessment model in comparison with VAR models.